In order to understand the distribution characteristics of meteorological disasters, realize the effective and accurate early warning of meteorological disasters, and reduce the losses caused by meteorological disasters to animal husbandry, a model for extracting the distribution characteristics of meteorological disasters in animal husbandry based on spectral clustering was proposed. By calculating the degree of impact, 11 meteorological disaster indicators are determined. Standardization, vacancy supplement and noise removal are carried out for meteorological disaster indicators. Spectral clustering algorithm is used to calculate the characteristic value of meteorological disasters, and the corresponding risk degree of meteorological disasters is obtained. The regions 1, 2, 5 and 7 show an upward trend, and the risk degree increases year by year, while the disaster characteristic values of regions 6 and 10 show a downward trend, and the risk degree decreases year by year; The disaster characteristic values of regions 3, 4, 8 and 9 have been maintained at a relatively stable level, of which the risk level of regions 3, 4 and 8 has been maintained at a low risk level, and the risk level of region 9 has been maintained at a moderate risk level. The low risk area occupies a larger area and is mostly concentrated in the northwest corner, the moderate risk area is mostly concentrated in the middle, and the high risk area is mostly concentrated in the southeast. The results show that the model has a good application effect and can effectively extract the distribution characteristics of meteorological disasters in animal husbandry.